The project will examine empirically the effect of minimum drinking age laws on traffic fatalities. The broad hypothesis to be tested is that "new drinking"--i.e., drinking by the inexperienced--creates a traffic fatality risk that is at least partially independent of the age at which drinking is legally permitted. This hypothesis is distinct from those examined in earlier investigations. If confirmed, it would call into question the use of minimum drinking age as a traffic safety tool. Confirmation would suggest that any reductions in fatalities among immediately-affected age groups when the minimum age is raised are in part offset by increased fatalities when the higher minimum age is reached. Thus, this policy would appear in part to postpone traffic fatalities, rather than to reduce them permanently.